Deep neural network application for 4D seismic inversion to changes in pressure and saturation: Optimizing the use of synthetic training datasets     

Gustavo Côrte  Jesper Dramsch  Hamed Amini  Colin MacBeth 《Geophysical Prospecting》2020,68(7):2164-2185

In this work, we tackle the challenge of quantitative estimation of reservoir dynamic property variations during a period of production, directly from four-dimensional seismic data in the amplitude domain. We employ a deep neural network to invert four-dimensional seismic amplitude maps to the simultaneous changes in pressure, water and gas saturations. The method is applied to a real field data case, where, as is common in such applications, the data measured at the wells are insufficient for properly training deep neural networks, thus, the network is trained on synthetic data. Training on synthetic data offers much freedom in designing a training dataset, therefore, it is important to understand the impact of the data distribution on the inversion results. To define the best way to construct a synthetic training dataset, we perform a study on four different approaches to populating the training set making remarks on data sizes, network generality and the impact of physics-based constraints. Using the results of a reservoir simulation model to populate our training datasets, we demonstrate the benefits of restricting training samples to fluid flow consistent combinations in the dynamic reservoir property domain. With this the network learns the physical correlations present in the training set, incorporating this information into the inference process, which allows it to make inferences on properties to which the seismic data are most uncertain. Additionally, we demonstrate the importance of applying regularization techniques such as adding noise to the synthetic data for training and show a possibility of estimating uncertainties in the inversion results by training multiple networks.  相似文献   

Inter- and intra-annual trend analysis of water level and flow in the middle and lower reaches of the Ganjiang River,China     

Di Zhu  Yue Ben  Xinfa Xu 《水文科学杂志》2020,65(12):2128-2141

ABSTRACT

The Ganjiang River is the largest tributary of Poyang Lake in China, and its hydrological regime variation greatly affects the utilization of regional water resources and the ecological environment of the lake. In this study, a novel trend analysis method, the Moving Average over Shifting Horizon (MASH), was applied to investigate the inter- and intra-annual trends of flow and water level from 1976 to 2016 at the Xiajiang and the Waizhou hydrological stations in the Ganjiang River. The Significant Change Rate Method (SCRM) was proposed to determine the MASH averaging parameters. The trend analysis results show a statistically significant decrease in water level series throughout the year and the relationship of flow and water level have changed greatly at the Waizhou station. The sediment load reduction, large-scale sand mining and water level decrease of Poyang Lake are identified as the main causes for the water level decrease.  相似文献   

基于分段副产物瞬时选择性的反应器网络综合     

胡仰栋  周勇  伍联营 《中国海洋大学学报(自然科学版)》2020,50(3):61-66

针对第三类反应体系的反应器网络综合问题,提出将串联反应以反应物到目的产物的反应步数为依据的新分段策略。通过比较不分段目的产物、不分段目的产物和中间产品、分段副产物和分段目的产物这四种不同的瞬时选择性,提出以分段副产物瞬时选择性为参数,以其对关键组分浓度或转化率的导数为依据选择各阶段最合适的反应器,以目的产物总选择性最大为目标采用遗传算法优化反应器网络。通过三个实例验证了本文所建立方法在解决复杂反应反应器网络综合问题的有效性,且本方法同样适用于简单反应体系。  相似文献   

基于PSR-WSVM模型的边坡位移预测     

李建新  刘小生  肖钢  周文  刘仁志 《大地测量与地球动力学》2020,40(6):577-580

为建立高精度的边坡位移预测模型，采用相空间重构（PSR）将边坡位移时间序列数据转换为多维数据，同时构造小波核函数改进的支持向量机模型，建立PSR-WSVM模型并应用于边坡位移预测。将PSR-WSVM模型预测结果与传统支持向量机（SVM）模型、小波支持向量机（WSVM）模型和基于相空间重构的支持向量机（PSR-SVM）模型预测结果进行对比，通过平均绝对误差（MAE）、平均绝对误差百分比（MAPE）和均方根误差（RMSE）3个精度评价指标验证PSR-WSVM模型的可行性。工程实例结果表明，PSR-WSVM模型预测结果的3个精度评价指标都优于另外3种模型，边坡位移预测的精度明显提升。  相似文献   

Redox evolution of western Tianshan subduction zone and its effect on deep carbon cycle     

《地学前缘(英文版)》2020,11(3):915-924

Knowing the phase relations of carbon-bearing phases at high-pressure(HP) and high-temperature(HT) condition is essential for understanding the deep carbon cycle in the subduction zones.In particular,the phase relation of carbon-bearing phases is also strongly influenced by redox condition of subduction zones,which is poorly explored.Here we summarized the phase relations of carbon-bearing phases(calcite,aragonite,dolomite,magnesite,graphite,hydrocarbon) in HP metamorphic rocks(marble,metapelite,eclogite) from the Western Tianshan subduction zone and high-pressure experiments.During prograde progress of subduction,carbonates in altered oceanic crust change from Ca-carbonate(calcite) to Ca,Mg-carbonate(dolomite),then finally to Mgcarbonate(magnesite) via Mg-Ca cation exchange reaction between silicate and carbonate,while calcite in sedimentary calcareous ooze on oceanic crust directly transfers to high-pressure aragonite in marble or amorphous CaCO3 in subduction zones.Redox evolution also plays a significant effect on the carbon speciation in the Western Tianshan subduction zone.The prograde oxygen fugacity of the Western Tianshan subduction zone was constrained by mineral assemblage of garnet-omphacite from FMQ-1.9 to FMQ-2.5 at its metamorphic peak(maximum P-T) conditions.In comparison with redox conditions of other subduction zones,Western Tianshan has the lowest oxygen fugacity.Graphite and light hydrocarbon inclusions were ubiqutously identified in Western Tianshan HP metamorphic rocks and speculated to be formed from reduction of Fe-carbonate at low redox condition,which is also confirmed by high-pressure experimental simulation.Based on petrological observation and high-pressure simulation,a polarized redox model of reducing slab but oxidizing mantle wedge in subduction zone is proposed,and its effect on deep carbon cycle in subduction zones is further discussed.  相似文献   

Deep gold mineralization features of Jiaojia metallogenic belt,Jiaodong gold Province: Based on the breakthrough of 3000 m exploration drilling     

《China Geology》2020,3(3):385-401

Recently, continuous breakthroughs have been made about deep gold prospecting in the Jiaodong gold province area of China. Approximately 5000 t of cumulative gold resources have been explored in Jiaodong, which has thus become an internationally noteworthy gold ore cluster. The gold exploration depth has been increased to about 2000 m from the previous <1000 m. To further explore the mineralization potential of the Jiaodong area at a depth of about 3000 m, the Shandong Institute of Geological Sciences has drilled an exploratory drillhole named “Deep drillhole ZK01” to a depth of 3266 m. Hence, as reported herein, the mineralization characteristics of the Jiaojia metallogenic belt have been successfully documented. ZK01 is, to date, the deepest borehole with an gold intersect in China, and constitutes a significant advance in deep gold prospecting in China. The findings of this study further indicate that the depth interval of 2000 m to 4000 m below the ground surface in the Wuyi Village area incorporates 912 t of inferred gold resources, while the depth interval of 2000 m to 4000 m below the surface across the Jiaodong area possesses about 4000 t of inferred gold resources. The Jiaojia Fault Belt tends to gently dip downward, having dip angles of about 25° and about 20° at vertical depths of 2000 m and 2850 m, respectively. The deep part of the Jiaojia metallogenic belt differs from the shallow and moderately deep parts about fracturing, alteration, mineralization, and tectonic type. The deep zones can generally be categorized from inside outward as cataclastic granite, granitic cataclasite, weakly beresitized granitic cataclasite, beresitized cataclasite, and gouge. These zones exhibit a gradual transitional relation or occur alternately and repeatedly. The mineralization degree of the pyritized cataclastic granite-type ore in the deep part of the Jiaojia metallogenic belt is closely related to the degree of pyrite vein development; that is, the higher the pyrite content, the wider the veins and the higher the gold grade. Compared to the shallow gold ores, the deep-seated gold ores have higher fineness and contain joseite, tetradymite, and native bismuth, suggesting that the deep gold mineralization temperature is higher and that mantle-sourced material may have contributed to this mineralization. ZK01 has also revealed that the deep-seated ore bodies in the Jiaojia metallogenic belt are principally situated above the main fracture plane (gouge) and hosted within the Linglong Granite, contradicting previous findings indicating that the moderately shallow gold ore bodies are usually hosted in the contact zone between the Linglong Granite and Jiaodong Group or meta-gabbro. These new discoveries are particularly significant because they can help correct mineralization prospecting models, determine favorable positions for deep prospecting, and improve metallogenic prediction and resource potential evaluation.  相似文献   

黔北正安地区五峰组—龙马溪组页岩气地质特征及有利区预测     

贺永忠  向坤鹏  安亚运  易成兴  杨忠琴  于宁 《中国地质调查》2020,7(3):21-29

为明确黔北正安地区上奥陶统五峰组至下志留统龙马溪组的页岩气地质特征,充分应用页岩气地质调查、重磁电、二维地震、地质调查井及测试分析等工作方法,开展黔北正安地区1:5万页岩气基础地质调查,对五峰组至龙马溪组富有机质页岩的分布、沉积环境、有机地球化学、岩石矿物、储集性能及含气性等特征进行分析研究,结果表明: 研究区五峰组—龙马溪组富有机质页岩为深水陆棚相沉积,主要分布有安场向斜、斑竹向斜和泥高向斜,厚度14.5~55 m,埋深0~3 200 m; 有机碳(TOC)含量1.0%~4.0%,有机质镜质体反射率(R_o)为1.82%~2.23%,有机质类型以Ⅰ型干酪根为主; 岩石主要由石英、长石和黏土矿物组成,脆性矿物含量高,一般大于50%; 孔隙度为2.03%~3.89%,渗透率为0.35×10^-5~1.86×10^-5μm²,表现为低孔、特低渗的特征; 最高含气量为2.88 m³/t,显示出较好的含气性特征。综合分析和评价圈出3个页岩气聚集有利区,分别为安场区块、斑竹区块和泥高区块,这为研究区页岩气进一步勘探开发指明了方向。  相似文献   

中国城市体系网络化研究   总被引：14，自引：1，他引：13  

胡国建  陈传明  金星星  王强 《地理学报》2019,74(4):681-693

当前,学者普遍认同城市体系研究的重心从等级范式向网络范式转向。在大量的文献梳理基础上,阐述20世纪80年代以来中国城市体系研究范式的转向过程,并进一步梳理中国城市网络研究的进展、问题与关注重点。结果表明：2010年之后,网络成为中国城市体系研究的主流范式,等级体系文献数量呈波动递减趋势,全国和长三角成为热点关注区域;属性数据一定程度上弥补了早期城市网络研究中数据匮乏的状况,关系数据应成为未来研究的侧重点;不同研究视角及其所应用的数据存在着一定的局限性,多元视角的综合研究成为趋势。在此基础上,提出城市网络研究未来可能的突破点,包括：调整国家城市网络的研究框架、超越交通数据直接考察城市间的实体联系、测度企业外部的横向联系、有向加权网络分析和可视化、增加研究中的地理“意味”等。  相似文献   

用改进的经验格林函数方法模拟唐山地震动     

席桂梅  郭雷  何书耕  张诗若  闵也  李启成 《震灾防御技术》2019,14(1):97-107

建筑物的抗震设防需要尽可能地掌握未来大地震强震动记录信息，但大地震强震动记录的匮乏阻碍了抗震设防实践的发展。经验格林函数方法作为模拟地震动的主要方法，可以提供可靠的大地震强震动记录，但也存在着许多问题，如缺乏对大地震断层滑动分布不均匀的描述、用经验确定小震数目、模拟方法受到大小地震相似条件的限制等。文中对上述经验格林函数方法存在的问题进行了研究，改进的经验格林函数方法，有效地解决了上述问题。并用其对唐山大地震进行了模拟，并把模拟的地震动时程和反应谱与实际记录相比较，发现用改进方法模拟的地震动加速度反应谱比用未改进方法模拟结果更接近实际的地震动记录加速度反应谱。由此说明改进的经验格林函数可更准确的模拟地震动。  相似文献   

Stripe-based fragility analysis of multispan concrete bridge classes using machine learning techniques     

Sujith Mangalathu  Jong-Su Jeon 《地震工程与结构动力学》2019,48(11):1238-1255

A framework for the generation of bridge-specific fragility curves utilizing the capabilities of machine learning and stripe-based approach is presented in this paper. The proposed methodology using random forests helps to generate or update fragility curves for a new set of input parameters with less computational effort and expensive resimulation. The methodology does not place any assumptions on the demand model of various components and helps to identify the relative importance of each uncertain variable in their seismic demand model. The methodology is demonstrated through the case study of a multispan concrete bridge class in California. Geometric, material, and structural uncertainties are accounted for in the generation of bridge numerical models and their fragility curves. It is also noted that the traditional lognormality assumption on the demand model leads to unrealistic fragility estimates. Fragility results obtained by the proposed methodology can be deployed in a risk assessment platform such as HAZUS for regional loss estimation.  相似文献